A

pcch

cd©

K

1

ptfbmc

ahtlDs

s

0d

Available online at www.sciencedirect.com

Journal of Pharmaceutical and Biomedical Analysis 45 (2007) 590–598

Compatibility of the antitumoral �-lapachone withdifferent solid dosage forms excipients

Marcılio S.S. Cunha-Filho, Ramon Martınez-Pacheco, Mariana Landın ∗Departamento de Farmacia y Tecnologıa Farmaceutica, Facultad de Farmacia,Universidad de Santiago de Compostela, Santiago de Compostela 15782, Spain

Received 7 May 2007; received in revised form 31 July 2007; accepted 6 August 2007Available online 19 August 2007

bstract

The objective of the present study was to evaluate the compatibility of the �-lapachone (�LAP), an antitumoral drug in clinical phase, withharmaceutical excipients of common use including diluents, binders, disintegrants, lubricants and solubilising agents. Differential scanningalorimetry (DSC) was used for a first screening to find small variations in peak temperatures and/or their associated enthalpy for six drug/excipientombinations (magnesium stearate, sodium estearyl fumarate, dicalcium phosphate dihydrate, mannitol, randomized methyl-�-cyclodextrin andydroxypropyl-�-cyclodextrin), which indicate some degree of interaction.

Additional studies using Fourier transformed infrared spectroscopy (FTIR), optical microscopy (OM) and heating–cooling DSC (HC-DSC)onfirmed the incompatibility of �LAP with magnesium stearate and dicalcium phosphate dihydrate. Those excipients should be avoided in theevelopment of solid dosage forms.

2007 Elsevier B.V. All rights reserved.

imetr

ptiadcl

csSaitt

eywords: �-Lapachone; Compatibility; Excipient; Differential scanning calor

. Introduction

The �-lapachone (�LAP) is an antitumoral drug in clinicalhase that presents good perspectives for its incorporation in theherapies in the near future. �LAP is obtained on a small scalerom South American trees of the families Bigoniaceae and Ver-enaceae [1]. On a larger scale it can be produced following theethod developed by Hooker and co-workers in 1892 through

yclization of lapachol in sulphuric acid [2].Recent studies have demonstrated that �LAP possesses

potent antineoplasic activity. In vitro and in vivo studiesave shown that �LAP inhibits conventional therapy-resistantumors, particularly the malignant neoplasm of slow cell cycleike prostate, colon and some ovarian and breast cancer [3–5].espite its interesting potential applications, preformulation

tudies have not yet been performed.The study of drug–excipient compatibility is an important

tage in the development of a solid dosage form as their incom-

∗ Corresponding author. Tel.: +34 981 563100x15044; fax: +34 981 549148.E-mail address: mlandin@usc.es (M. Landın).

aiac

do

731-7085/$ – see front matter © 2007 Elsevier B.V. All rights reserved.oi:10.1016/j.jpba.2007.08.016

y

atibility can alter the stability and/or the biovailability of drugs,hereby, affecting its safety and/or efficacy. Two types of chem-cal incompatibilities have been described between excipientsnd drugs: those corresponding to intrinsic chemical drug degra-ation such hydrolysis or oxidation but without significant directovalent chemical reactions, and those corresponding to cova-ent reaction between the drug and the excipient [6].

Throughout the different methods reported on drug–excipientompatibility studies, DSC has been shown to be a rapid,ensitive and simple technique used in routine experiments.mall variations in peak temperature or associated enthalpyre interpreted as an indication of interaction and possiblencompatibility. Despite the advantages of DSC there are cer-ain limitations. The extrapolation of findings obtained at highemperatures, are not always in agreement with the drug real situ-tion in the formulation [7,8]. Therefore, the DSC results must benterpreted carefully and some complementary techniques, suchs infrared spectroscopy, microscopy or X-ray diffractometry

an be useful in avoiding misleading conclusions [9].

In the present study the compatibility of �LAP with 12ifferent pharmaceutical excipients of common use in the devel-pment of solid dosage forms was evaluated. Five cyclodextrins

eutica

hhb

2

2

pwMpts(hGdM(FccRmcb(

2

�srowm

2

wfab

hsd

s1fca

2

pamowDf

2

sgpae

2

ycT

3

3

o

eaatesite

mtltheie

M.S.S. Cunha-Filho et al. / Journal of Pharmac

ave also been included because �LAP has an extremely lowydrosolubility and, recently, cyclodextrins have been shown toe useful in overcoming this problem [10–12].

. Experimental

.1. Materials

�LAP was a gift from UFPE, Brazil (lot L503) with aurity of 99.9% by HPLC and DSC. Following excipientsere purchased from commercial sources and used as such.icrocrystalline cellulose PH 101 (MCC, Guinama/Spain),

regelatinized maize starch (starch, Guinama/Spain), lac-ose monohydrated (lactose, Guinama/Spain), magnesiumtearate (MGST, Guinama/Spain), sodium estearyl fumaratePRUV, Julia-Parrera/Spain), talc (talc, Guinama/Spain),ydroxypropylmethyl cellulose K100LV Premium (HPMC,uinama/Spain), polyvinyl pyrrolidone (PVP, BASF/Germany),icalcium phosphate dihydrate Emcompress (DCPD, Unitedendell/UK), medium-viscosity carboxymethyl cellulose

CMC, Sigma/Espana), sodium croscarmellose (explocel,MC/USA), mannitol (mannitol, AstraZeneca/Spain). Theyclodextrins �-cyclodextrin (�CD) and randomized methyl-�-yclodextrin (RM�CD, molar substitution 0.5) were donated byoquette/France; the hydroxypropyl-�-cyclodextrin (HP�CD,olar substitution 2.7) was facilitated by Janssen Pharma-

eutica/Belgium; the �-cyclodextrin (�CD) was providedy Wacker/Germany and the sulfobutylether-�-cyclodextrinSB�CD, molar substitution 1.0) was donate by Cydex/EUA.

.2. Study protocol

The flow diagram (Fig. 1) presents the protocol used for theLAP–excipient compatibility evaluation. Firstly, DSC mea-urements on binary �LAP–excipient mixtures were selected forapid screening. Investigation on mixtures showing interactionsr changes in the thermal profile of components were completedith Fourier transformed infrared spectroscopy (FTIR), opticalicroscopy (OM) and heating–cooling DSC studies (HC-DSC).

.3. Sample preparation

�LAP–excipient physical mixtures in a ratio of 1:1 (w/w)ere prepared by mixing in a Turbula WAB T2C (Switzerland)

or 15 min. This proportion was chosen to maximize the prob-bility of interactions between materials. Those samples haveeen denoted as PM.

PM samples were stored in an oven at 40 ◦C and 75% relativeumidity in closed containers using sodium chloride saturatedolutions for a month [13]. Those stressed samples have beenenoted as SM.

Additionally, the drug, some excipients and its mixtures wereubjected to thermal stress by heating up to 160 ◦C at a rate of

0 ◦C min−1. This treatment was proven to be non-destructiveor the isolated materials, which maintained their physical andhemical entity when cooling. Those samples have been denoteds TSM.

pacw

l and Biomedical Analysis 45 (2007) 590–598 591

.4. Differential scanning calorimetry (DSC)

Samples weighing 3–4 mg were placed in open aluminiumans and heated from 30 to 300 ◦C at a rate of 10 ◦C min−1 usingtemperature modulated DSC Q100 calorimeter (TA Instru-ents, USA). Nitrogen was used as purge gas at a flux rate

f 50 mL min−1. The calibration of temperature and heat flowas performed with standard indium samples. Heating–coolingSC (HC-DSC) studies were carried out at a rate of 10 ◦C min−1

ollowing the sequence 30–170–30–300 ◦C.

.5. Fourier transformation-infrared spectroscopy (FTIR)

Infrared spectra were obtained using a Bruker IFS-66Vpectrometer (Bruker Daltonics Inc., Germany). Samples wereround, mixed thoroughly with potassium bromide, and com-ressed in a hydraulic press. Thirty-two scans were obtainedt a resolution 4 cm−1. Results were compared with referencexcipient spectra in the literature [14].

.6. Optical microscopy (OM)

The morphological characteristics of the samples were anal-sed using an Olympus SZ60 (Opelco, Japan) microscopeonnected to a video camera Olympus DP12 (Opelco, Japan).he images were processed using Analysis® version 3.2.

. Results and discussion

.1. Drug–excipient compatibility first DSC screening

DSC data of the �LAP and excipient thermal events in singler PM and SM binary systems are presented in Tables 1 and 2.

The �LAP DSC curve (Table 1) displays two events; a sharpndotherm at 157.3 ◦C, due to the drug melting with an associ-ted enthalpy of 104.0 ± 1.5 J g−1 and the drug decompositionfter 230 ◦C. This profile is characteristic of an anhydrous crys-alline substance. The melting �LAP event is highly repeatableven when �LAP is blended with the different excipients (peakhifts higher than 0.5 ◦C are in bold type in Table 1). Variationsn the enthalpy values for the binary mixtures can be attributedo some heterogeneity in the small samples used for the DSCxperiments (3–4 mg) [7,9].

Some excipient endothermic events are melting (lactose,annitol, MGST or PRUV) or dehydration phenomena (lac-

ose, DCPD) but the most are broad peaks associated to theiross of adsorbed water (Table 2) with inherent important varia-ions in their enthalpies, particularly for the SM samples whichave been stored at 40 ◦C and 75% RH. A careful analysis ofxcipient thermal behaviour is required in order to avoid mis-nterpretations. Bold type indicates important changes in thexcipient DSC profiles in Table 2.

Thermograms obtained by DSC analysis of individual com-

onents were compared with those of the mixtures, before (PM)nd after its storage at 40 ◦C and 75% RH (SM). Of the 17ommonly used excipients tested only six showed interactionsith �LAP. The profiles of the mixtures were an overlapping of

592 M.S.S. Cunha-Filho et al. / Journal of Pharmaceutical and Biomedical Analysis 45 (2007) 590–598

P–excipient compatibility study.

ttw[iDtah

cDw1[s

Fig. 1. Strategy for the �LA

he thermograms of the single components for MCC, starch, lac-ose, talc, HPMC, PVP, CMC, explocel, �CD, �CD and SB�CD,hich is accepted as compatibility evidence between materials

15]. Changes in �LAP events and/or excipient DSC profilesn the binary systems (PM) could be detected in two diluents,CPD and mannitol; two lubricants, MGST and PRUV and

wo solubilising agents HP�CD and RM�CD. Those variationsre slightly accentuated with ageing at 40 ◦C and 75% relativeumidity (SM samples).

Mixtures of �LAP with mannitol and DCPD have shownhanges, among diluents studied, in both drug and excipientsSC events (Figs. 2 and 3). Mannitol is an excipient with a

ell-defined thermal profile characterized by a melting peak at68.9 ◦C indicative by its crystalline and anhydrous state (Fig. 2)16]. Binary blending �LAP–mannitol (PM and SM) presenthifts in melting peaks of drug and mannitol, as have been Fig. 2. DSC analysis of �LAP, mannitol and �LAP–mannitol combinations.

M.S.S. Cunha-Filho et al. / Journal of Pharmaceutical and Biomedical Analysis 45 (2007) 590–598 593

Table 1Thermal data from DSC of �LAP melting event, in single and PM and SM binary systems studied

po

odta(

emoi

ointed out for other drugs such as amoxicillin, paracetamolr vitamin D [17–19].

The DCPD DSC profile is associated with the loss of waterf hydration from this material and the formation of anhy-

rous dicalcium phosphate. The dehydration process occurs inwo stages at 150.8 and 191.1 ◦C (associated enthalpies 100nd 310 J g−1, respectively) in agreement with different authorsTable 2), the contribution of which depends on the DCPD vari-

Fig. 3. DSC analysis of �LAP, DCPD and �LAP–DCPD combinations.

(tr

tTosilappwmiNsia(c

ty and its particle size [20,21]. As it can be observed (Fig. 3),ixing �LAP with DCPD caused a shift in the melting peak

f the drug and the disappearance of the DCPD character-stic events. Additionally, the enthalpy associated 155.5 J g−1

Table 1), is lower than that corresponding to the addition ofhe thermal phenomena of components (about 250 J g−1). Theseesults suggest a robust interaction between components.

The MGST has been shown to be incompatible with an impor-ant number of drugs, such as glibenclamide or aspirin [22,23].his excipient presents two molecules of water hydration (5.5%f weight approximately). The dehydration process occurs ineveral steps at over 70 ◦C. Additionally, it presents a melt-ng peak at approximately 120 ◦C, which sometimes appears atower temperatures or even superimposes the dehydration eventss a consequence of the product being a mixture of magnesiumalmitate and stearate [9,24,25]. Our DSC results agree withrevious information (Fig. 4 and Table 2), MGST showing aide endotherm with two events at 73.0 and 103.7 ◦C. The PMixture thermogram presents a slight shift in the �LAP melt-

ng peak and a reduction in the associated enthalpy (Table 1).o changes can be denoted in the MGST events. However, the

tressed sample (SM), exhibits important modifications regard-

ng the MGST thermogram (Table 2). Those variations can bettributed to the modifications associated to the MGST ageingMGSTaged) at the above conditions (40 ◦C and 75% RH) as itan be seen in the MGSTaged DSC profile (Fig. 4). The hydration

594 M.S.S. Cunha-Filho et al. / Journal of Pharmaceutical and Biomedical Analysis 45 (2007) 590–598

Table 2Thermal data from DSC of excipients studied and PM and SM binary systems

oc

iTeira2i

mb

loiw

f MGST seems to increase the interaction with �LAP, whichould affect on the drug stability negatively.

Mixtures of �LAP and PRUV also exhibit diverse variationsn the thermal behaviour of both products (Fig. 5 and Table 2).he PRUV DSC profile is characterized by three endothermicvents at 110, 135 and 200 ◦C [26]. The PM mixtures showmportant variations in the positions of the three events. In this

◦

espect, peaking at 110 C it seems to split into two; the peakt 135 ◦C shifts to a lower temperature (131 ◦C) and the one at00 ◦C disappears completely (Table 2). Additionally, a widen-ng in drug melting peak can be observed. Those differences are

Fig. 4. DSC analysis of �LAP, MGST and �LAP–MGST combinations.

btoit

aintained in the SM samples. The slight hygroscopic nature ofoth, �LAP and PRUV, justifies this behaviour [16].

Cyclodextrins (CDs) are cyclic oligosaccharides withipophilic inner cavities and hydrophilic outer surfaces capablef interacting with a large variety of drugs giving non-covalentnclusion complexes [27]. �LAP can form inclusion complexesith different cyclodextrins giving an improvement in its solu-ility and dissolution rate [10,12]. Variations in the raw material

hermal profiles or even the disappearance of the fusion peakf the drug in combinations of drug–cyclodextrin are oftennterpreted as an evidence of an inclusion complex forma-ion. Cyclodextrins exhibit a broad endothermic effect ranging

Fig. 5. DSC analysis of �LAP, PRUV and �LAP–PRUV combinations.

M.S.S. Cunha-Filho et al. / Journal of Pharmaceutical and Biomedical Analysis 45 (2007) 590–598 595

Fig. 6. DSC analysis of �LAP, RM�CD, �LAP–RM�CD combinations. Fig. 7. DSC analysis of �LAP, HP�CD, �LAP–HP�CD combinations.

Fig. 8. OM micrographs of some of the mixtures indicated.

596 M.S.S. Cunha-Filho et al. / Journal of Pharmaceutical and Biomedical Analysis 45 (2007) 590–598

F . Banda

bcatatitsstecb�

3a

cccHcOibtfbip

oam

cmcbifa[

dci�t�LAP crystals and also coloured excipient particles (Fig. 8).

The heating–cooling DSC results suggest that �LAP pro-motes DCPD dehydration at lower temperature, at the same timeas the melted drug (Fig. 12). The DCPD water hydration (20.9%

ig. 9. FTIR spectra and data of �LAP, HP�CD and its mixtures SM and TSMnd b (broad).

etween 30 and 149 ◦C associated with water loss from inside theavity (Table 2). Among the cyclodextrins studied, only HP�CDnd RM�CD (Figs. 6 and 7) shows small modifications in thehermal profiles of the stressed samples (SM), mainly shiftingt the melting peak of the drug and in the enthalpies associatedo loss of water from the CDs, which could be indicative of thenteractions. The �LAP–RM�CD mixtures (Fig. 6) show, addi-ionally, an extra exothermic event at 178 ◦C, which was reallymall quantitatively (associated enthalpy 8 J g−1) but was con-istent. This could be associated to a strong interaction betweenhe RM�CD and the melted drug and subsequently, the �LAPntrance into the empty CD cavity in an energetic favoured pro-ess that decreases the energy of the system. The RM�CD haseen shown as the most useful �CD derivative in improvingLAP solubility [12].

.2. FTIR, OM and HC-DSC compatibility confirmatoryssays

Neither mannitol nor PRUV incompatibility with �LAP wasonfirmed from those additional experiments. No significanthanges in the FTIR characteristic bands of the pure substancesan be detected, even after heating at 160 ◦C (TSM samples).C-DSC experiments show, on cooling, two exothermal events

orresponding to the drug and excipient crystallization process.n the second heating, the endothermal events of the first heat-

ng remain unchanged. No sign of decomposition was detectedy optical microscopy (as an example see mannitol/�LAP mix-ures, Fig. 8). Results are in agreement with other author findingsor mannitol [7,9,19] and PRUV [26]. The interaction detectedy thermal methods can be explained on the basis of the proxim-ty of the melting events of both components without significanthysicochemical stability problems.

The same conclusion was achieved after additional researchn RM�CD and HP�CD/�LAP samples. FTIR data (Fig. 9)nd HC-DSC (Fig. 10) on HP�CD/�LAP and RM�CD/�LAPixtures, respectively do not show significant changes on the

s are classified in function of its intensity as s (strong), m (medium), w (weak)

haracteristic behaviour of the pure substances even after ther-al stress (TSM) or after a melting–crystallization–melting

ycle. Interactions between those CDs and the drug suggestedy the preliminary DSC results (Table 1) should be interpretedn terms of the inclusion of �LAP into the CD cavity and theormation of inclusion complexes, thus being a beneficial inter-ction with profitable repercussion in the �LAP hydrosolubility12].

The DCPD/�LAP incompatibility was confirmed by FTIRata (Fig. 11). Accentuated changes were detected in the bandsorresponding to the functional groups of the drug, especiallyn the TSM samples, which indicate chemical decomposition ofLAP. The OM results showed the intrinsic association between

he compounds. The mixture (TSM) is composed of several

Fig. 10. HC-DSC of the �LAP–RM�CD PM binary system.

M.S.S. Cunha-Filho et al. / Journal of Pharmaceutical and Biomedical Analysis 45 (2007) 590–598 597

Fig. 11. FTIR spectra and data of �LAP, DCPD and its mixtures SM and TSM. Bandand b (broad).

Fig. 12. HC-DSC of the �LAP–DCPD PM binary system.

oeastsmswtrcrd

F(t

Fig. 13. FTIR spectra and data of �LAP, MGST and its mixtures SM and TSM. Bandand b (broad).

s are classified in function of its intensity as s (strong), m (medium), w (weak)

f the DCPD weight) partially dissolves the �LAP and the basicnvironment from this [28] could decompose the �LAP, whichre especially sensitive to an alkaline pH. This was verified on theecond heating, by a broad drug melting peak, shifting to loweremperature and with a slight associated enthalpy, thus a clearign of drug degradation. The DCPD is one of the most com-on diluents for tablets but it has been described as an excipient

usceptible to dehydration at low temperature in the presence ofater vapour, which is clearly relevant to the choice of condi-

ions for processing and storage of the dosage forms [20]. Ouresults suggest great modifications in both �LAP and DCPDharacteristics as a result of their incompatibility that would notecommend the use of DCPD in the development of �LAP solidosage forms.

The MGST/�LAP interactions have also been confirmed by

TIR data (Fig. 13). The disappearance of different MGST bands1579.5 and 1465.7 cm−1) from both, SM and TSM FTIR spec-ra, suggests the decomposition of the excipient. Additionally,

s are classified in function of its intensity as s (strong), m (medium), w (weak)

598 M.S.S. Cunha-Filho et al. / Journal of Pharmaceutica

sarcoaemtorwHt

4

amuimef

A

J

tFc(pA

R

[

[

[

[[

[

[

[[

[

[

[[

[

[

[

Fig. 14. HC-DSC of the �LAP–MGST PM binary system.

ome modifications in the �LAP bands (2978 and 1568 cm−1)ssigned to the aromatic region of the drug, C H and C C,espectively, take place. The HC-DSC experiments (Fig. 14)orroborate the hypothesis of MGST degradation as in the sec-nd heating only the melting peak of �LAP, at lower temperaturend associated enthalpy, can be observed and none of the MGSTvents. Heating up to 160 ◦C does not promote changes in theorphology of �LAP or MGST single systems but when heated

ogether (Fig. 8) the sample becomes black as a consequencef the degradation of MGST. Lubricants are used in tablets at aatio far lower than the one used in this study so, at the ratio 1:1/w greater MGST degradation would be expected on ageing.owever, the interaction evidences between products suggest

hat MGST should be avoided in the �LAP formulations.

. Conclusions

Six (magnesium stearate, PRUV, DCPD, mannitol, RM�CDnd HP�CD) of the seventeen excipients studied presented ther-al interactions with the �LAP. However, additional studies

sing FTIR, optical microscopy and heating–cooling DSC stud-es (HC-DSC) confirmed the incompatibility of �LAP withagnesium stearate and dicalcium phosphate dihydrate. Those

xcipients should be avoided in the development of solid dosageorms.

cknowledgements

Authors thank LAFEPE/Brazil, and Professor Dr. Pedroose Rolim Neto, Federal University of Pernambuco/Brazil, for

[

[[

l and Biomedical Analysis 45 (2007) 590–598

heir kind gift of the �LAP. Janssen Pharmaceutica, Roquettereres and CyDex are grateful for the generous donation ofyclodextrins. This work was supported by Xunta de GaliciaPGIDIT05BTF20301PR) and the Programme Al�an, the Euro-ean Union Programme of High level Scholarships for Latinmerica, scholarship no. E04D043994BR.

eferences

[1] A.R. Burnett, R.H. Thomson, J. Chem. Soc. C 7 (1968) 850–853.[2] S.C. Hooker, H.W. Shepard, J.G. Walsh Jr., G.H. Connitt, J. Am. Chem.

Soc. 58 (1936) 90–1197.[3] C.J. Li, C. Wang, A.B. Pardee, Cancer Res. 55 (1995) 3712–3715.[4] M. Planchon, S. Wuerzberger, B. Frydman, D.T. Witiak, P. Hutson, D.R.

Church, G. Wilding, D.A. Boothman, Cancer Res. 55 (1995) 3706–3711.[5] M. Ough, A. Lewis, E.A. Bey, J. Gao, J.M. Ritchie, W. Bornmann, D.A.

Boothman, L.W. Oberley, J.J. Cullen, Cancer. Biol. Ther. 4 (2005) 95–102.

[6] G. Damien, STP Pharma Pratiques 14 (2004) 303–310.[7] R.K. Verma, S. Garg, J. Pharm. Biomed. Anal. 38 (2005) 633–644.[8] D. Kiss, R. Zelko, C. Novak, Z. Ehen, Therm. Anal. Calorim. 84 (2006)

447–451.[9] P.C. Mora, M. Cirri, P. Mura, J. Pharm. Biomed. Anal. 42 (2006) 3–10.10] N. Nasongkla, A.F. Wiedmann, A. Bruening, M. Beman, D. Ray, W.G.

Bornmann, D.A. Boothman, J. Gao, Pharm. Res. 20 (2003) 1626–1633.11] F. Wang, E. Blanco, H. Ai, D.A. Boothman, J. Gao, J. Pharm. Sci. 95 (2006)

2309–2319.12] M. Cunha-Filho, B. Dacunha-Marinho, J.J. Torres-Labandeira, R.

Martınez-Pacheco, M. Landın, AAPS Pharm. Sci. Technol. 8 (2007),doi:10.1208/pt0803060, Article 60.

13] H. Nyqvist, Int. J. Pharm. Technol. Prod. 4 (1983) 47–48.14] D.E. Bugay, W.P. Findlay, Pharmaceuticasl Excipients: By IR, Raman, and

NMR Spectroscopy, Marcel Dekker, New York, 1999, 669 p.15] J. Wells, Pharmaceutical Preformulation, Ellis Horwood, Chichester, 1988,

227 p.16] American Pharmaceutical Association, Handbook of Pharmaceutical

Excipients, 4th ed., Pharmaceutical Press, London, 2006, 918 p.17] J. Sawicka, Pharmazie 46 (1991) 519–521.18] M.A. Hassan, J. Kaloustian, P. Prinderre, H. Ramsis, K.A. Khaled, T.H.

El-Faham, S.S. Tous, L. Maury, J. Joachim, Pharmazie 51 (1996) 400–403.

19] M. Tomassetti, A. Catalani, V. Rossi, S. Vecchio, J. Pharm. Biomed. Anal.37 (2005) 949–955.

20] M. Landın, R. Martınez-Pacheco, J.L. Gomez-Amoza, C. Souto, A.Concheiro, R.C. Rowe, Int. J. Pharm. 103 (1994) 9–18.

21] M. Landın, R.C. Rowe, P. York, Int. J. Pharm. 104 (1994) 271–275.22] G. Pyramides, J.W. Robinson, S.W. Zito, J. Pharm. Biomed. Anal. 13 (1995)

103–110.23] G.G. Oliveira, H.G. Ferraz, J.S.R. Matos, Therm. Anal. Calorim. 79 (2005)

267–270.24] A. Marini, V. Berbenni, S. Moioli, G. Bruni, P. Cofrancesco, C. Margheritis,

M. Villa, Therm. Anal. Calorim. 73 (2003) 529–545.25] R.K. Verma, S. Garg, J. Pharm. Biomed. Anal. 35 (2004) 449–458.

26] C.E. Malan, M.M. Villiers, A.P. Lotter, J. Pharm. Biomed. Anal. 15 (1997)

549–557.27] M.E. Davis, M.E. Brewster, Nat. Rev. Drug Discov. 12 (2004) 1023–1035.28] M. Landın, M.J. Fontao, R. Martinez-Pacheco, Drug Dev. Ind. Pharm. 31

(2005) 249–256.